540 



REVISION OP TliE AXINELMDiE, 11., 



of their length gradually (except frequently for slight irregu- 

 larities near the apex) to a sharp point; a gradual slight narrow- 

 ing of the spicule towards the basal end, however, is not uncom- 

 mon. The curvature as a rule is restricted to the basal moiety 

 of the spicule, and is usually well-pronounced, but varies much 



both in form and degree : 

 frequently it is more or less 

 angulate, the spicule ap- 

 pearing slightly bent; and 

 occasional spicules are bi- 

 angulate. In odd cases of 

 extreme curvature, the 

 form of the spicule makes 

 some approach to that of 

 a rhabdostyle. Variability 

 exists also in the shape of 

 the spicule at its basal ex- 

 tremity, which frequently 

 shows a tendency to be- 

 come abruptly more or less 

 sharp-pointed, either has- 

 tately or mucronately so; 

 but sometimes the attenu- 

 ation is more gradual, and 

 the form assumed is that 

 of an oxea. The propor- 

 tion of oxeote forms is 

 greatest amongst the slen- 

 derer, presumably imma- 

 ture spicules, which only 

 occur scattered between tlie fibres, and are relatively few in 

 number. The maximal size of the spicules in the case of 

 Dendy's specimens is given as 290 x 166/x; in the present speci- 

 men, the size attained is 350 x 15/x, but individuals much ex- 

 ceeding 320/x in length are scarce; those composing the fibres 

 are seldom less in stoutness than IO/7.. Developmental forms of 

 all sizes down to less than 1 40 x l/x are to be met with. (In the 



Text fig. 15. — SI(/i)i(Lvla Jiabe/Zala. 

 a, nie<^asclei'es of tiie fibres; //, intei- 

 stitial niegascleres; f, signiata. 



